Isobaric Vapor–Liquid Equilibrium of Binary Systems of Hexane or

Apr 3, 2012 - nonrandom two-liquid (NRTL), and the universal quasichemical (UNIQUAC) ... isobaric VLE data were measured for the binary systems of...
2 downloads 0 Views 858KB Size
Article pubs.acs.org/jced

Isobaric Vapor−Liquid Equilibrium of Binary Systems of Hexane or Octane with 1,2-Dimethylbenzene or 1,3-Dimethylbenzene at 101.3 kPa B. S. Gupta and Ming-Jer Lee* Department of Chemical Engineering, National Taiwan University of Science and Technology, 43 Keelung Road, Section 4, Taipei 106-07, Taiwan ABSTRACT: Isobaric vapor−liquid equilibrium (VLE) data for the binary mixtures of hexane or octane with 1,2-dimethylbenzene or 1,3-dimethylbenzene were obtained at 101.3 kPa by using a recirculation type of apparatus. All of the experimental VLE data passed the thermodynamic consistency tests of the Herington and the van Ness methods. No azeotrope was found in these binary systems, and each system exhibited large deviations from ideal behavior. The experimental VLE data were correlated well with the Wilson, the nonrandom two-liquid (NRTL), and the universal quasichemical (UNIQUAC) activity coefficient models, respectively. The predictive capability of the universal functional (UNIFAC) and the conductor-like screening model for realistic solvents (COSMO-RS) was also evaluated with the experimental values. These two models fail to quantitatively predict the VLE properties for the octane-containing systems.



INTRODUCTION The knowledge of vapor−liquid equilibrium (VLE) properties of mixtures is not only fundamentally but also practically important for process development. In the refinery industry, the separation of aliphatic and aromatic hydrocarbons is commonly encountered. VLE data of the related mixtures are essential. Recently, Calvarb et al.1 found that dimethylbenzenes can be extracted from a liquid mixture of dimethylbenzenes and hexane by using the ionic liquid 1-ethyl-3-methylimidazolium ethylsulfate, [EMim][ESO4]. Also, several studies on other thermodynamic properties at different conditions have been published.2−4 For the system of hexane + 1,4-dimethylbenzene, isothermal VLE data at 313.15 K have been reported by Goral,5 isobaric VLE data (T−x) for the same system at 95 kPa by Prasad et al.,6 and also isobaric VLE data (T-x,y) at 101.3 kPa by Michishita et al.7 The isobaric VLE data for the binary systems of 1,2-, 1,3-, or 1,4-dimethylbenzene with cyclohexane and heptane at 101.3 kPa were reported by Tojo et al.,8,9 with methyl-2propanol at 101.3 kPa by Monton et al.,10 with ethenylbenzene at (5 and 15) kPa by Loras et al.,11 with ethylbenzene at 100.65 kPa by Mattedi et al.,12 and with ethylbenzene at (6.66 and 26.66) kPa by Monton et al.13 In our previous work,14 we reported the isobaric VLE data of nonane with cyclohexane, methylbenzene, 1,3-dimethylbenzene, or 1,4-dimethylbenzene at 101.3 kPa. In this paper, the isobaric VLE data were measured for the binary systems of hexane or octane with 1,2-dimethylbenzene or 1,3-dimethylbenzene at 101.3 kPa by using a circulating type of apparatus. The Herington15 area test and the point-to-point test of the van Ness16 method as modified by Fredunsland et al.17 were used to check the thermodynamic consistency of the experimental © 2012 American Chemical Society

data. In the reduction of VLE data, the fugacity coefficient of each constituent compound in the vapor phase was calculated by using the two-term virial equation of state incorporated with the Hayden and O'Connell (HOC) model18 for estimating the second virial coefficients. The nonideality of liquid phase was represented by the Wilson,19 the nonrandom two-liquid (NRTL),20 or the universal quasichemical (UNIQUAC)21 activity coefficient models. In this study, we also checked the predictive capability of the universal functional (UNIFAC)17 and the conductor-like screening model for realistic solvents (COSMO-RS)22 for the systems investigated.



EXPERIMENTAL SECTION Materials. All of the chemicals used were of analytical grade and purchased from commercial sources. The suppliers and the purities of the chemicals are given in Table 1. The purity of each chemical was checked by gas chromatography (GC) and the impurity levels found were less than 0.003 in mass fraction. The purities of the chemicals were reconfirmed by measuring their density (ρ) and boiling point, and compared with literature values. Both showed good agreement with the literature values,23−26 as can be also seen in Table 1. The density of pure components was measured by using a digital vibrating-tube densimeter (DMA 4500, Anton Paar), with an accuracy of ± 0.00005 g·cm−3. Boiling points were measured by a recirculation-type equilibrium apparatus, which was equipped Received: December 19, 2011 Accepted: March 22, 2012 Published: April 3, 2012 1237

dx.doi.org/10.1021/je3002283 | J. Chem. Eng. Data 2012, 57, 1237−1243

Journal of Chemical & Engineering Data

Article

Table 1. Material Description, Densities (ρ) at T = 298.2 K, and Boiling Points T of Pure Componentsa ρ/g·cm−3 compound

source

hexane octane 1,2-dimethylbenzene 1,3-dimethylbenzene

mass fraction purity

Acros Aldrich Alfa Aesar Acros

purification method

0.9972 0.9994 0.9995 0.9999

none none none none

this work 0.65529 0.69868 0.87550 0.86007

T/K lit.

this work

lit.

341.5 398.8 417.3 412.3

341.88b 398.86c 417.55d 412.27e

b

0.65493 0.69868c 0.8755d 0.86009e

a u(ρ) = 0.00005 g·cm−3; u(T) = 0.1 K, where u is the standard uncertainty. bTaken from ref 23. cTaken from ref 24. dTaken from ref 25. eTaken from ref 26.

Table 2. Experimental Vapor−Liquid Equilibrium Data for the Binary System of Hexane (1) + 1,2-Dimethylbenzene (2) at 101.3 kPaa

a

T/K

x1

y1

γ1

341.5 344.4 347.8 351.4 355.7 359.6 364.6 369.5 379.9 385.5 394.0 402.6 414.6 417.3

1.000 0.936 0.819 0.725 0.625 0.515 0.429 0.343 0.231 0.187 0.122 0.074 0.018 0.000

1.000 0.990 0.973 0.956 0.935 0.910 0.875 0.836 0.737 0.667 0.534 0.393 0.113 0.000

1.000 0.982 1.001 1.005 1.014 1.081 1.099 1.164 1.196 1.182 1.214 1.243 1.177

γ2

B11

B22

B12

1.656 1.378 1.283 1.178 1.091 1.074 1.031 1.000 1.003 1.006 0.972 0.971 1.000

−1371.7 −1339.2 −1302.6 −1265.5 −1223.3 −1186.9 −1142.5 −1101.5 −1021.5 −981.9 −926.1 −874.1 −808.4 −794.6

−2799.8 −2718.7 −2628.3 −2537.7 −2435.9 −2349.1 −2244.8 −2149.6 −1967.8 −1879.8 −1757.8 −1646.6 −1509.1 −1480.6

−703.1 −687.4 −669.5 −651.4 −630.5 −612.3 −589.9 −569.0 −527.6 −506.8 −477.2 −449.3 −413.4 −405.7

u(T) = 0.1 K; u(x1) = 0.005; u(y1) = 0.005.

Table 3. Experimental Vapor−Liquid Equilibrium Data for the Binary System of Hexane (1) + 1,3-Dimethylbenzene (2) at 101.3 kPaa

a

T/K

x1

y1

γ1

341.5 344.5 346.9 348.9 351.1 355.3 366.3 371.2 375.8 388.3 396.4 401.7 412.2

1.000 0.900 0.812 0.740 0.673 0.557 0.352 0.296 0.249 0.148 0.095 0.063 0.000

1.000 0.983 0.974 0.963 0.952 0.932 0.846 0.798 0.744 0.566 0.413 0.295 0.000

1.000 1.011 1.037 1.063 1.087 1.147 1.241 1.236 1.230 1.194 1.149 1.115

γ2

B11

B22

B12

1.499 1.109 1.055 1.000 0.893 0.933 0.955 0.975 0.988 0.994 0.995 1.000

−1371.7 −1338.1 −1312.1 −1291.1 −1268.5 −1227.1 −1128.0 −1087.8 −1051.9 −963.0 −911.1 −879.4 −821.0

−2727.9 −2646.8 −2584.7 −2534.7 −2481.4 −2384.6 −2157.7 −2067.6 −1988.3 −1795.6 −1686.1 −1620.0 −1500.5

−696.1 −679.9 −667.3 −657.0 −645.9 −625.4 −575.6 −555.1 −536.6 −490.0 −462.4 −445.3 −413.5

u(T) = 0.1 K; u(x1) = 0.005; u(y1) = 0.005.

with a thermocouple calibrated to an uncertainty of ± 0.1 K. All of the materials were used without further purification. Apparatus and Procedure. The isobaric VLE data were obtained by using a recirculation-type phase equilibrium still (NGW Co.). The schematic diagram and the working procedure of this apparatus were described by Shiah et al.27 In each experimental run, about 100 cm3 of liquid mixture was injected in the still. During the course of measurement, the isobaric condition (i.e., 101.3 ± 0.2 KPa) was maintained by using a pressure adjustment system directly attached to the equilibrium still.14 A mercury barometer (model 453, stability = ± 0.1 kPa, Princo Instruments) was used for atmospheric pressure measurement.

When the temperature remained constant for at least two hours, equilibrium conditions were assumed to be attained, and the equilibrium temperature was recorded. The equilibrium temperature was measured by a digital thermometer (TES 1310 type-K), which was calibrated with a precise platinum thermometer (model-1560, Hart Scientific Co.) to an uncertainty of ± 0.1 K. Vapor and liquid samples were collected for analysis under equilibrium conditions. Analysis. The compositions of vapor and liquid samples were analyzed by a gas chromatograph (model 8700, China Chromatography) with a thermal conductivity detector and a stainless-steel column (model SE-30, Supelcoport, 80/100 mess, 14′ × 1/8″, 20 %). Helium (purity > 0.9995 in mass 1238

dx.doi.org/10.1021/je3002283 | J. Chem. Eng. Data 2012, 57, 1237−1243

Journal of Chemical & Engineering Data

Article

Table 4. Experimental Vapor−Liquid Equilibrium Data for the Binary System of Octane (1) + 1,2-Dimethylbenzene (2) at 101.3 kPaa

a

T/K

x1

y1

γ1

398.8 399.4 401.0 402.6 404.5 406.0 408.8 409.3 410.5 412.3 415.0 415.7 417.3

1.000 0.924 0.786 0.667 0.537 0.436 0.309 0.297 0.234 0.182 0.091 0.065 0.000

1.000 0.942 0.836 0.740 0.641 0.557 0.438 0.423 0.360 0.291 0.163 0.123 0.000

1.000 1.004 1.005 1.006 1.032 1.064 1.101 1.093 1.147 1.141 1.199 1.246

γ2

B11

B22

B12

1.237 1.187 1.157 1.090 1.060 1.018 1.014 0.999 0.989 0.980 0.980 1.000

−1831.7 −1823.3 −1801.0 −1779.1 −1753.7 −1734.0 −1698.0 −1691.7 −1676.8 −1654.7 −1622.3 −1614.1 −1595.5

−1694.3 −1686.6 −1666.4 −1646.6 −1623.6 −1605.7 −1573.2 −1567.5 −1554.0 −1534.0 −1504.8 −1497.4 −1480.6

−632.3 −629.6 −622.4 −615.4 −607.1 −600.7 −588.9 −586.8 −581.8 −574.5 −563.7 −561.0 −554.7

u(T) = 0.1 K; u(x1) = 0.005; u(y1) = 0.005.

Table 5. Experimental Vapor−Liquid Equilibrium Data for the Binary System of Octane (1) + 1,3-Dimethylbenzene (2) at 101.3 kPaa

a

T/K

x1

y1

γ1

398.8 399.2 400.0 400.9 401.6 402.5 403.9 405.5 406.2 407.3 409.8 411.4 412.3

1.000 0.934 0.843 0.741 0.659 0.556 0.450 0.351 0.320 0.249 0.131 0.038 0.000

1.000 0.948 0.869 0.783 0.718 0.632 0.540 0.438 0.413 0.338 0.191 0.064 0.000

1.000 1.005 1.000 1.001 1.014 1.034 1.053 1.052 1.069 1.094 1.105 1.228

γ2

B11

B22

B12

1.116 1.156 1.132 1.096 1.072 1.041 1.033 1.011 1.002 0.991 0.994 1.000

−1831.7 −1826.1 −1814.9 −1802.4 −1792.8 −1780.5 −1761.7 −1740.5 −1731.4 −1717.1 −1685.5 −1665.7 −1654.7

−1655.6 −1650.6 −1640.7 −1629.7 −1621.2 −1610.4 −1593.7 −1575.1 −1567.0 −1554.5 −1526.6 −1509.1 −1499.4

−624.0 −622.2 −618.6 −614.6 −611.5 −607.6 −601.5 −594.6 −591.6 −587.0 −576.5 −570.0 −566.4

u(T) = 0.1 K; u(x1) = 0.005; u(y1) = 0.005.

fraction) was used as a carrier gas. The response of gas chromatography was analyzed by a peak-ABC chromatograph data handling system. Before making the analysis, the gas chromatography was calibrated by using several standard samples with known compositions over the entire composition range. These standard samples were prepared gravimetrically with an electronic balance (R&D model GR-200) to the uncertainty of ± 0.01 mg. By using the calibration curve, compositions of vapor and liquid phases are measured from the average of at least three replications. The uncertainty of the composition measurement was estimated to within ± 0.005 in mole fraction.



RESULTS AND DISCUSSION Experimental Data and Reliability Test. The isobaric VLE data (T, xi, yi) and the corresponding activity coefficient (γi) at 101.3 kPa for the binary systems of hexane (1) + 1,2dimethylbenzene (2), hexane (1) + 1,3-dimethylbenzene (2), octane (1) + 1,2-dimethylbenzene (2), and octane (1) + 1,3dimethylbenzene (2) are listed in Tables 2 to 5, respectively. The VLE phase diagrams of these binary systems are presented in Figures 1 to 4. All binary systems investigated were found to be nonazeotropic in nature. The thermodynamic consistency of all of the experimental data were checked by using the Herington area test method.15

Figure 1. Plot of experimental equilibrium temperature at 101.3 kPa against the mole fraction of hexane (1) + 1,2-dimethylbenzene (2); ●, experimental liquid phase and ○, experimental vapor phase; , calculated liquid and vapor compositions from the NRTL model; ----, calculated liquid and vapor compositions from the Wilson model; ···, calculated liquid and vapor compositions from the UNIQUAC model; -·-, predicted liquid and vapor compositions from the UNIFAC model; -··-, predicted liquid and vapor compositions from the COSMO-RS model. 1239

dx.doi.org/10.1021/je3002283 | J. Chem. Eng. Data 2012, 57, 1237−1243

Journal of Chemical & Engineering Data

Article

Figure 4. Plot of experimental equilibrium temperature at 101.3 kPa against the mole fraction of octane (1) + 1,3-dimethylbenzene (2); ●, experimental liquid phase and ○, experimental vapor phase; , calculated liquid and vapor compositions from the NRTL model; ----, calculated liquid and vapor compositions from the Wilson model; ···, calculated liquid and vapor compositions from the UNIQUAC model; -·-, predicted liquid and vapor compositions from the UNIFAC model; -··-, predicted liquid and vapor compositions from the COSMO-RS model.

Figure 2. Plot of experimental equilibrium temperature at 101.3 kPa against the mole fraction of hexane (1) + 1,3-dimethylbenzene (2); ●, experimental liquid phase and ○, experimental vapor phase; , calculated liquid and vapor compositions from the NRTL model; ----, calculated liquid and vapor compositions from the Wilson model; ···, calculated liquid and vapor compositions from the UNIQUAC model; -·-, predicted liquid and vapor compositions from the UNIFAC model; -··-, predicted liquid and vapor compositions from the COSMO-RS model.

Table 6. Results of the Herington Thermodynamic Consistency Test system

D−J

hexane + 1,2-dimethylbenzene hexane + 1,3-dimethylbenzene octane + 1,2-dimethylbenzene octane + 1,3-dimethylbenzene

0.5 3.2 5.3 0.5

Vapor−Liquid Equilibrium Calculation. At VLE, the fugacity of each component i in the vapor phase should equal to that in the liquid phase, fiV = fiL

(1)

yi ϕiV P = xi γi f iοL

(2)

or

where f, ϕ, y, P, x, and γ refer to the fugacity, fugacity coefficient, vapor composition, pressure, liquid composition, and activity coefficient, respectively. The superscripts V, L, and o represent the vapor phase, liquid phase, and standard state, respectively, and the subscript i stands for component i. From the above equation, the activity coefficient of component i can be expressed as

Figure 3. Plot of experimental equilibrium temperature at 101.3 kPa against the mole fraction of octane (1) + 1,2-dimethylbenzene (2); ●, experimental liquid phase and ○, experimental vapor phase; , calculated liquid and vapor compositions from the NRTL model; ----, calculated liquid and vapor compositions from the Wilson model; ···, calculated liquid and vapor compositions from the UNIQUAC model; -·-, predicted liquid and vapor compositions from the UNIFAC model; -··-, predicted liquid and vapor compositions from the COSMO-RS model.

γi = (yi ϕiV P)/{xiPisϕis exp[ViL(P − Pis)/RT ]}

(3)

where V, R, and T are the molar volume, gas constant, and temperature, respectively. The superscript s represents a saturation condition. The fugacity coefficients ϕiV and ϕis were calculated from the two-term virial equation by using the Hayden and O'Connell (HOC) model18 for the estimation of the self and the cross second virial coefficients (B11, B22, and B12). Since no complex formation in the systems was investigated, the association and salvation parameters (η)

The data can be considered to be thermodynamically consistent, if D − J < 10. Table 6 lists the results of the consistency test which shows that all of the systems pass this criterion. 1240

dx.doi.org/10.1021/je3002283 | J. Chem. Eng. Data 2012, 57, 1237−1243

Journal of Chemical & Engineering Data

Article

Table 7. Physical Properties of Pure Componentsa TC

PC

VC

μ

Rd

compound

K

kPa

cm3·mol−1

Debye

Å

ZC

ω

rb

qc

hexane octane 1,2-dimethylbenzene 1,3-dimethylbenzene

507.6 568.7 630.3 617.0

3025 2490 3732 3541

371 486 370 375

0 0 0.6296 0.2997

3.769 4.546 3.836 3.937

0.266 0.256 0.264 0.259

0.3013 0.3995 0.3101 0.3265

4.4997 5.8484 4.6579 4.6579

3.856 4.936 3.536 3.536

a Taken from the Aspen properties databank. model.

b

Volume parameter of the UNIQUAC model. cSurface area parameter of the UNIQUAC

Table 8. Parameters of the Extended Antoine Equationa,b compound

A1b

B2b

C3b

D4b

E5b

106·F6b

G7b

T1/K

T2/K

hexane octane 1,2-dimethylbenzene 1,3-dimethylbenzene

97.74 89.17 83.49 78.19

−6995.5 −7900.2 −7955.2 −7615.9

0 0 0 0

0 0 0 0

−2.70 −11.00 −10.08 −9.30

12.30 7.18 5.95 5.56

2 2 2 2

177.83 216.38 247.98 255.30

507.60 568.70 630.30 617.00

a

Taken from the Aspen property databank. bExtended Antoine equation: ln(Pis = A1 + (B2/T + C3) + D4T + E5 ln T + F6TG7 for T1 < T < T2, where Ps is in kPa and T in K.

Table 9. Interaction Binary Parameters and Root-Mean-Square Deviations model Wilsonb NRTLc,d UNIQUACe

Wilsonb NRTLc,d UNIQUACe

Wilsonb NRTLc,d UNIQUACe

Wilsonb NRTLc,d UNIQUACe

parameters M12 = 8.53 M21 = −7.58 A12 = 9.44 A21 = −11.08 P12 = −3.68 P21 = 4.86 M12 = 2.84 M21 = −14.81 A12 = 18.10 A21 = −0.17 P12 = −10.69 P21 = 1.89 M12 = 4.81 M21 = 0.17 A12 = −1.46 A21 = −5.52 P12 = −1.92 P21 = 4.23 M12 = 8.79 M21 = −0.98 A12 = 2.78 A21 = −11.47 P12 = −0.57 P21 = 2.39

rmsd ΔTa/K

rmsd ΔPa/kPa

Hexane (1) + 1,2-Dimethylbenzene (2) N12(K) = −3264.65 0.194 N21(K) = 2778.76 B12(K) = −3511.54 0.193 B21(K) = 4254.55 Q12(K) = 1439.18 0.191 Q21(K) = −1907.02 Hexane (1) + 1,3-Dimethylbenzene (2) N12(K) = −799.46 0.040 N21(K) = 4691.03 B12(K) = −5470.73 0.044 B21(K) = −380.25 Q12(K) = 3249.18 0.038 Q21(K) = −379.08 Octane (1) + 1,2-Dimethylbenzene (2) N12(K) = −2073.23 0.166 N21(K) = −44.45 B12(K) = 411.72 0.150 B21(K) = 2560.44 Q12(K) = 875.41 0.132 Q21(K) = −1867.41 Octane (1) + 1,3-Dimethylbenzene (2) N12(K) = −3761.66 0.112 N21(K) = 506.30 B12(K) = −1313.65 0.107 B21(K) = 4938.34 Q12(K) = 317.79 0.112 Q21(K) = −1073.17

rmsd 100Δxa

rmsd 100Δya

0.021

0.606

0.626

0.021

0.608

0.620

0.021

0.640

0.615

0.004

0.298

0.634

0.004

0.208

0.278

0.003

0.203

0.346

0.021

0.379

0.504

0.019

0.430

0.468

0.017

0.460

0.447

0.014

0.222

0.227

0.013

0.227

0.226

0.014

0.232

0.229

np rmsd ΔM = (1/np∑k=1 (Mkcalc − Mkexpt)2)0.5, where np is the number of data points and M represents T, P, x1, or y1. bWilson model: ln Aij = [Mij + c Nij/T]. NRTL model: τij = [Aij + Bij/T]. dThe value of α was fixed at 0.3 for each binary system. eUNIQUAC model: τij = [exp(Pij + Qij/T)]. a

vapor pressure (Pis) of component i was calculated from the extended Antoine equation:

were assigned as zero for all components. The input parameters, including the critical properties (TC and PC), dipole moment (μ), and the radius of gyration (Rd), are given in Table 7. The calculated values of virial coefficients for each system investigated are listed in Tables 2 to 5. The tabulated activity coefficients (γi) in Tables 2 to 5 were calculated from eq 3 with the experimental values of xi, yi, T, and P, and the

ln(Pis) = A1 +

B2 + D4 T + E5 ln T + F6T G7 T + C3

(4)

The values of the coefficients A1 to G7 for each component i were taken from the Aspen property databank and are listed in 1241

dx.doi.org/10.1021/je3002283 | J. Chem. Eng. Data 2012, 57, 1237−1243

Journal of Chemical & Engineering Data

Article

the binary systems of hexane or octane with 1,2-dimethylbenzene or 1,3-dimethylbenzene at 101.3 kPa. In general, the NRTL model yields slightly better results. According to the thermodynamic consistency test method of van Ness16 modified by Fredunsland et al.,17 the experimental VLE data are consistently provided, and the mean absolute deviation between calculated and measured vapor composition, Δy, is less than 0.01. As can be seen from Table 9, all of the binary systems satisfied this criterion. In this study we also attempt to test the predictive capability of a group contribution model, the UNIFAC,17 and a quantum approach based method, the COSMO-RS,22 for the systems investigated. The comparison between experimental and predicted values can be seen in Figures 1 to 4. Both the UNIFAC and the COSMO-RS models fail to quantitatively predict the VLE properties of the octane-containing systems (as shown in Figures 3 and 4), while good agreement between the predicted and the experimental values was found from the hexane-containing systems (as shown in Figures 1 and 2). To understand the nonideality of these binary systems investigated, the excess Gibbs free energies GE were calculated by using the following equation:

Figure 5. Plot of estimated values of excess Gibbs free energy at 101.3 kPa from the NRTL model against mole fraction; , hexane (1) + 1,2-dimethylbenzene (2); ----, hexane (1) + 1,3-dimethylbenzene (2); -·-, octane (1) + 1,2-dimethylbenzene (2); ···, octane (1) + 1,3dimethylbenzene (2).

G E = RT (x1 ln γ1 + x2 ln γ2)

where the activity coefficients, γi, were estimated from the NRTL model using the binary parameters as given in Table 9. Figure 5 shows the variation of excess Gibbs free energy with composition for these four systems. Each system investigated exhibits positive values of GE over the entire composition range. The values of GE for these systems follow the order of hexane (1) + 1,2-xylene (2) > hexane (1) + 1,3-xylene (2) > octane (1) + 1,2-dimethylbenzene (2) > octane (1) + 1,3-dimethylbenzene (2).

Table 8. The liquid molar volume ViL of pure component i was estimated from the Rackett model28 modified by Yamada and Gunn.29 The critical properties (TC, PC, VC, ZC), acentric factor (ω), dipole moment (μ), the radius of gyration (Rd), and the parameters of the UNIQUAC model (r and q) of each component, which are needed in the phase equilibrium calculation, were also taken from Aspen property databank and are listed in Table 7. Correlation and Prediction. The Wilson,19 NRTL,20 and UNIQUAC21 combined with the HOC models were used to correlate the experimental VLE data for each binary system. The value of αij in the NRTL model was fixed at 0.3. The data correlation was performed by using Aspen plus V7.1. In the data reduction process, the Britt and Luecke30 algorithm based on the maximum likelihood principle was adopted to minimization of the following objective function: np ⎧⎡



CONCLUSION Isobaric VLE data have been determined experimentally for the binary systems of hexane + 1,2-dimethylbenzene, hexane + 1,3dimethylbenzene, octane + 1,2-dimethylbenzene, and octane + 1,3-dimethylbenzene at 101.3 kPa over the entire composition range. No azeotrope formation was observed these four binary systems. In general, the nonideality of these four systems follows the order of hexane (1) + 1,2-xylene (2) > hexane (1) + 1,3-xylene (2) > octane (1) + 1,2-dimethylbenzene (2) > octane (1) + 1,3-dimethylbenzene (2). The Herington area test and the point-to-point test of the van Ness as modified by Fredunsland et al. were used to check the reliability of experimental results, and all four binary systems passed both thermodynamic consistency test criteria. The Wilson-HOC, the NRTL-HOC, and the UNIQUAC-HOC models were used, respectively, to correlate the experimental isobaric VLE data for each binary system. Good correlations were found from all three models, with the NRTL model yielding slightly better results. These new VLE data were also used to examine the predictive capability of two solution models, UNIFAC and COSMO-RS. It was found that both models failed to quantitatively predict the VLE properties for octane-containing systems.

2 expt ⎤

⎪⎢ (P calc − Pk F = ∑ ⎨⎢ k σp k=1 ⎪ ⎩⎣

2 ⎡ calc expt ⎤ − )⎥ ( T T ) ⎢ k ⎥ k ⎥ +⎢ ⎥ σT ⎦ ⎣ ⎦

⎡ (x calc − x expt) ⎤2 ⎡ (y calc − y expt ) ⎤2 ⎫ 1, k ⎥ 1, k ⎥ ⎪ ⎢ 1, k ⎢ 1, k +⎢ +⎢ ⎥ ⎥ ⎬ σx1 σy ⎣ ⎦ ⎣ ⎦ ⎪ ⎭ 1

(6)

(5)

where np is the number of data points. The superscripts of calc and expt represent the calculated and the experimental values, respectively. The standard deviations (σ) of temperature and pressure are 0.1 K and 0.2 kPa, respectively, and those of liquid and vapor compositions are 0.005. The results of data correlation are given in Table 9, and graphical representation can be seen in Figures 1 to 4. Figures 1 to 4 reveal that the calculated results from these three models almost overlap with the experimental points. Table 9 also shows that the deviations between the experimental and the calculated values are very small. All of these three models represent satisfactorily the VLE behavior for



AUTHOR INFORMATION

Corresponding Author

*Tel.: +886-2-2737-6626; fax: +886-2-2737-6644. E-mail address: [email protected] (M.J.L.). 1242

dx.doi.org/10.1021/je3002283 | J. Chem. Eng. Data 2012, 57, 1237−1243

Journal of Chemical & Engineering Data

Article

Funding

(17) Fredenslund, A.; Gmehling, J.; Rasmussen, P. Vapor-Liquid Equilibria Using UNIFAC; Elsevier: Amsterdam, 1977. (18) Hayden, J. G.; O'Connell, J. P. A Generalized Method for Predicting Second Virial Coefficients. Ind. Eng. Chem. Process Des. Dev. 1975, 14, 209−216. (19) Wilson, G. M. Vapor-Liquid Equilibrium XI. A New Expression for the Excess Free Energy of Mixing. J. Am. Chem. Soc. 1964, 86, 127−130. (20) Renon, H.; Prausnitz, J. M. Local Composition in Thermodynamics Excess Function for Liquid Mixtures. AIChE J. 1968, 14, 135− 144. (21) Abrams, D. S.; Prausnitz, J. M. Statistical Thermodynamics of Liquid Mixtures: A New Expression for the Excess Energy of Partially or Completely Miscible Systems. AIChE J. 1975, 21, 116−128. (22) Klamt, A.; Eckert, F.; Diedenhofen, M. Prediction or Partition Coefficients and Activity Coefficients of Two Branched Compounds Using COSMOtherm. Fluid Phase Equilib. 2009, 285, 15−18. (23) Dominguez, M.; Cea, P.; Lopez, M. C.; Royo, F.; Urieta, J. S. Isobaric VLE Data of the Binary Mixture (n-Hexane + 1Chlorobutane) and the Ternary System (1-Butanol + n-Hexane + 1Chlorobutane) at 101.3 kPa. Fluid Phase Equilib. 1999, 164, 195−207. (24) Hiaki, T.; Taniguchi, A.; Tsuji, T.; Hongo, M. Isothermal Vapor Liquid Equilibria of Octane with 1-Butanol, 2-Butanol, or 2-Methyl-2Propanol. Fluid Phase Equilib. 1998, 144, 145−155. (25) Diaz, C.; Tojo, J. Isobaric Vapor-Liquid Equilibria of the Binary Systems Cyclohexane with o-Xylene, m-Xylene, p-Xylene, and Ethylbenzene at 101.3 kPa. J. Chem. Eng. Data 2002, 47, 1154−1158. (26) TRC Thermodynamic Tables; Thermodynamics Research Center, Texas A&M University: College Station, TX, 1996. (27) Shiah, I. M.; Yau, T. S.; Tseng, H. C. Vapor-Liquid Equilibrium of Water + Acetic Acid, and Water + Propionic Acid Systems Saturated with Sodium Chloride. J. Chin. Inst. Chem. Eng. 2000, 31, 41−47. (28) Rackett, H. G. Equation of State for Saturated Liquids. J. Chem. Eng. Data 1970, 15, 514−517. (29) Yamada, T.; Gunn, R. D. Saturated Liquid Molar Volumes. Rackett Equation. J. Chem. Eng. Data 1973, 18, 234−236. (30) Britt, H. I.; Luecke, R. H. The Estimation of Parameters in Nonlinear Implicit Models. Technometrics 1973, 15, 233−247.

The authors gratefully acknowledge the financial support from the National Science Council, Taiwan, through Grant No. NSC99-2221-E-011-079-MY3 and scholarship from National Taiwan University of Science & Technology. Notes

The authors declare no competing financial interest.

■ ■

ACKNOWLEDGMENTS The authors thank Dr. Ho-mu Lin for valuable discussions. REFERENCES

(1) Gonzalez, E. J.; Gonzalez, B.; Calvarb, N.; Dominguez, A. Study of [EMim][ESO4] Ionic Liquid as Solvent in the Liquid-Liquid Extraction of Xylenes from Their Mixtures with Hexane. Fluid Phase Equilib. 2011, 305, 227−232. (2) Exarchos, N. C.; Margari, M. T.; Demetropoulos, I. N. Viscosities and Densities of Dilute Solutions of Glycerol Trioleate + Octane, + p-Xylene, + Toluene, and + Chloroform. J. Chem. Eng. Data 1995, 40, 567−571. (3) Yang, C.; Xu, W.; Ma, P. Thermodynamic Properties of Binary Mixtures of p-Xylene with Cyclohexane, Heptane, Octane, and N-Methyl-2-pyrrolidone at Several Temperatures. J. Chem. Eng. Data 2004, 49, 1794−1801. (4) Jain, D. V. S.; Chadha, R.; Sehgal, S. K. Excess Molar Volumes of Binary Mixtures Containing Ethylbenzene and Isomeric Xylenes at 303.65, 312.65 and 321.65 K. Fluid Phase Equilib. 1995, 112, 151−156. (5) Goral, M. Vapour-Liquid Equilibria in Non-polar Mixtures. III. Binary Mixtures of Alkylbenzenes and n-Alkane at 313.15 K. Fluid Phase Equilib. 1994, 102, 275−286. (6) Prasad, T. E. V.; Prasad, D. H. L.; Patanjali, V. Bubble Temperature Measurements on Seven Binary Mixtures Containing 1−4-Dimethylbenzene at 95.8 kPa. Chem. Eng. Commun. 2005, 192, 18−22. (7) Michishita, T.; Arai, Y.; Saito, S. Vapor-Liquid Equilibria of Hydrocarbons at Atmospheric Pressure. Kagaku Kogaku 1971, 35, 111−116. (8) Diaz, C.; Tojo, J. Isobaric Vapor-Liquid Equilibria of the Binary Systems Cyclohexane with o-Xylene, m-Xylene, p-Xylene, and Ethylbenzene at 101.3 kPa. J. Chem. Eng. Data 2002, 47, 1154−1158. (9) Diaz, C.; Tojo, J. Phase Equilibria Behaviour of n-Heptane with o-Xylene, m-Xylene, p-Xylene and Ethylbenzene at 101.3 kPa. J. Chem. Thermodyn. 2002, 34, 1975−1984. (10) Lores, M. T.; Torre, D. L.; Burguet, M. C.; Monton, J. B. VaporLiquid Equilibria for the Binary Systems Isobutanol with m-Xylene, o-Xylene and p-Xylene at 101.3 kPa. J. Chem. Eng. Data 1999, 44, 869−872. (11) Aucejo, A.; Loras, S.; Martinez-Soria, V.; Becht, N.; Rio, G. D. Isobaric Vapor-Liquid Equilibria for the Binary Mixtures of Styrene with Ethylbenzene, o-Xylene, m-Xylene, and p-Xylene. J. Chem. Eng. Data 2006, 51, 1051−1055. (12) Rodrigues, W. L.; Mattedi, S.; Abreu, J. C. N. Vapor-Liquid Equilibria Data for Binary Systems of Ethylbenzene + Xylene Isomers at 100.65 kPa. J. Chem. Eng. Data 2005, 50, 1134−1138. (13) Monton, J. B.; Llopis, F. J. Isobaric Vapor-Liquid Equilibria of Ethylbenzene + m-Xylene and Ethylbenzene + o-Xylene Systems at 6.66 and 26.66 kPa. J. Chem. Eng. Data 1994, 39, 50−52. (14) Gupta, B. S.; Lee, M. J. Isobaric Vapor-Liquid Equilibrium for the Binary Mixtures of Nonane with Cyclohexane, Toluene, m-Xylene or p-Xylene at 101.3 kPa. Fluid Phase Equilib. 2011, 313, 190−195. (15) Herington, E. F. G. Tests for the Consistency of Experimental Isobaric Vapor-Liquid Equilibrium Data. J. Inst. Pet. 1951, 37, 457− 470. (16) Van Ness, H. C.; Byer, S. M.; Gibbs, R. E. Vapor-Liquid Equilibrium. Part I. An Appraisal for Data Reduction Methods. AIChE J. 1973, 19, 238−244. 1243

dx.doi.org/10.1021/je3002283 | J. Chem. Eng. Data 2012, 57, 1237−1243